Smart Instrumentation Retrofit for Continuous Process Monitoring

The water treatment facility processed approximately 50,000 cubic meters per day through a sequence of filtration, chemical dosing, and disinfection stages. Despite the scale and criticality of the operation, process monitoring relied heavily on manual gauge readings taken by operators every four hours. Between readings, the plant was effectively operating without real-time visibility into key parameters.

This gap had measurable consequences. Chemical dosing—particularly chlorine injection—was based on scheduled volume rather than real-time demand, leading to periodic over-dosing that wasted chemicals and under-dosing that risked compliance failures. Pump cavitation events were only detected after operators noticed abnormal noise or vibration, by which point bearing damage had already occurred. Filter differential pressure was checked manually using analog gauges, meaning filter backwash cycles were time-based rather than condition-based, resulting in both premature and delayed backwash events.

The facility had a legacy PLC system (Siemens S7-300) with available I/O capacity, but no field instruments were connected to it. The client wanted a sensor network that could be integrated into the existing control system without replacing the PLC, and that would provide continuous data to both the local SCADA system and a future cloud-based analytics platform.

The engineering team designed the instrumentation network in three functional layers: field sensing, signal conditioning and communication, and integration with the existing PLC.

At the field level, 34 sensor points were installed across the treatment process. Pressure transmitters (4–20mA output, 0–10 bar range) were placed at pump discharge points, filter inlets, and filter outlets to enable real-time differential pressure calculation. RTD PT100 temperature sensors were installed at chemical storage tanks and at post-disinfection monitoring points. Electromagnetic flow meters were fitted on main trunk lines and on chemical dosing lines to provide volumetric flow data with ±0.5% accuracy. Ultrasonic level transmitters were installed on raw water intake tanks and on chemical holding tanks to replace the existing float-based level switches, which had a failure rate exceeding 15% annually.

IFM proximity sensors were installed on all rotary equipment—pumps, blowers, and mixers—to detect shaft rotation and enable run-hour tracking. These sensors used IO-Link protocol, which allowed both discrete and diagnostic data to be transmitted on the same cable.

The communication layer used IFM IO-Link masters connected via industrial Ethernet to a signal gateway that converted IO-Link data into PROFINET packets compatible with the existing Siemens PLC. Analog sensors (4–20mA) were connected through signal conditioners that provided galvanic isolation and converted the analog signals to digital IO-Link format before reaching the gateway. This architecture eliminated the need for dedicated analog input cards on the PLC, using instead the PLC's existing PROFINET port.

The SCADA system was updated with new display screens showing real-time process data, trend graphs, and alarm setpoints configured for each measured parameter.

Within the first 60 days of operation, the instrumentation network fundamentally changed how the facility was managed. Operators transitioned from four-hourly manual rounds to continuous SCADA monitoring with exception-based intervention.

Chemical dosing optimization was the most immediate impact. By linking chlorine injection rates to real-time flow measurements, the facility reduced chlorine consumption by approximately 18% while maintaining compliance at all sampling points. Filter backwash cycles were switched from time-based to condition-based triggers using differential pressure thresholds, reducing backwash water consumption by roughly 25% and extending average filter run times by four hours.

Pump health monitoring via proximity sensors and pressure transmitters enabled the maintenance team to identify two developing cavitation issues before bearing failure occurred. In both cases, the issue was traced to partially closed suction valves that had been inadvertently left in a throttled position. Catching these issues early avoided an estimated AED 45,000 in repair costs and associated downtime.

Nuisance alarms—previously averaging 40 per shift due to unreliable float switches and manual reading errors—dropped by approximately 70% once the new transmitters provided stable, accurate signals. This allowed operators to trust the alarm system and respond to genuine process deviations rather than dismissing alerts as false positives.

Overall plant uptime improved from approximately 91% to 96% over the first six months, with the majority of the improvement attributable to reduced unplanned pump outages and optimized filter management.

This project illustrates a pattern common across process industries in the region: facilities with capable control systems but insufficient field instrumentation to leverage them. The existing PLC had the processing power and I/O capacity to manage sophisticated control loops, but without sensors providing continuous data, it was functioning as little more than a motor starter panel.

The IO-Link architecture proved particularly well-suited for retrofit applications because it minimized cabling requirements and allowed both simple discrete sensors and complex analog transmitters to communicate on the same network infrastructure. This reduced installation time and cost compared to conventional point-to-point 4–20mA wiring.

For facilities considering similar instrumentation upgrades, the key takeaway is that sensor deployment should be designed around specific operational decisions: what parameter must be measured, at what frequency, with what accuracy, and how will the data drive a control action or maintenance decision. Sensors installed without a clear decision framework generate data without generating value.

Creative Automation is currently working with this client on a Phase 2 deployment that extends the IO-Link network to the sludge handling and dewatering systems, and introduces cloud-based data logging for long-term trend analysis and regulatory reporting.